1. Field of the Invention
The present invention relates to an analog-signal processing circuit provided with switches for carrying out averaging of signals.
2. Description of the Related Art
In a conventional CMOS image sensor, a large number of pixels for outputting voltages corresponding to quantities of light are arranged in order to acquire image information. Outputs from these pixels are converted into a color image by being subjected to A/D conversion and digital signal processing. Here, the higher the number of pixels, the finer the optical information can be acquired, and hence, in recent years, a product in which a large number of pixels are arranged in one chip has become the mainstream.
However, when the number of pixels becomes large, the resolution of the image is increased, whereas there are a drawback that the time necessary for A/D conversion and digital signal processing is increased due to an increase in the data amount, and the time required to take an image is prolonged, and a drawback that the number of images that can be stored in a recording medium the recording capacity of which is limited is reduced. For this reason, a CMOS image sensor of recent years is equipped with a function of making the number of signals to be subjected to A/D conversion less than the actual number of pixels (see, for example, U.S. Pat. No. 6,794,627). However, an arithmetic mean with a high degree of accuracy has not been obtained by the conventional averaging method.
An analog signal processing circuit capable of improving the accuracy of an arithmetic mean of signals is required.